Janus kinases (Jaks) have broad roles in immune regulation via their action in cytokine signalling [1–3]. These non-receptor tyrosine kinases phosphorylate receptor chains, which in turn recruit and phosphorylate members of the Signal Transducer and Activator of Transcription (STAT) family [2, 4]. The Jak family comprises Jak1, Jak2, Jak3 and Tyk2. These enzymes have very similar domain structures, containing a FERM domain, an SH2 domain, a pseudokinase domain, and a catalytic tyrosine kinase domain. Jaks serve overlapping but distinct functions in cytokine signaling, as demonstrated by knockout, mutation and other studies [5–9].
Because of their roles in the signaling of many important cytokines, hormones, and growth factors such as IL-2, IL-4, IL-6, IL-7, IL-12, IL-13, IFN-α, IFN-γ, Epo, and GM-CSF [10, 11], Jak inhibitors might have wide application in the treatment of inflammatory, myeloproliferative and autoimmune diseases, and therefore the Jak enzymes are attractive targets for drug discovery. Initial studies with Jak3 inhibitors were aimed at preventing solid organ transplant rejection [12, 13]. More recent studies have explored the potential of such compounds in chronic autoimmune diseases such as rheumatoid arthritis and psoriasis [14–16]. For example, tofacitinib (CP-690,550), which inhibits Jak1, Jak2, and Jak3, has demonstrated efficacy in Phase II trials for rheumatoid arthritis [17–19]. Ruxolitinib (Jakafi®), a dual Jak1 and Jak2 inhibitor , was recently approved for the treatment of myelofibrosis, a disorder involving myeloproliferative neoplasm.
The development of Tyk2 inhibitors is less advanced. Tyk2 functions together with Jak2 in the signaling of IL-12 and IL-23 via its interaction with the IL-12Rβ1 receptor chain, and in the coordinated phosphorylation of STAT3 & STAT4 [4, 21]. Human Tyk2 gene deficiency causes defects in signaling of multiple cytokines, including IL-6, IL-10, IL-12 and IL-23, and reduced production of IFNγ . Furthermore, Tyk2-deficient mice are resistant to experimental autoimmune encephalomyelitis, a model for multiple sclerosis [22, 23]. Given the importance of Tyk2-dependent downstream cytokine signaling in this and other diseases such as rheumatoid arthritis and Crohn’s disease, Tyk2 inhibitors have the potential to be important therapeutics.
Because Jak family active sites exhibit high sequence identity, designing inhibitors selective within the family is challenging. One way to approach this challenge is to target active site regions that differ in conformation between homologs. To identify these “hot-spot” regions, we set out to obtain multiple crystal structures of Tyk2 in complex with a variety of ligands representing diverse chemotypes. At the time of our initial work, only Jak2 and Jak3 crystal structures had been published [24, 25]. Robust Tyk2 crystallography allowing for the soaking of multiple inhibitors, essential for rapid throughput in structure-based drug design, had not been described. After exploring multiple constructs, we obtained crystals of mouse Tyk2 in the presence of 3-aminoindazole inhibitors that diffracted to 2.5–2.6 Å resolution. The inclusion of a ligand was absolutely required to obtain high-quality crystals, and we found through limited proteolysis experiments that the enzyme is significantly stabilized by binding to such ATP-competitive inhibitors. This process enabled the determination of multiple inhibitor-soaked Tyk2 crystal structures, forming the basis of an extensive SBDD program.